Science - USA (2022-04-29)

genotypes and, in some cases, help explain the
overall cellular phenotypes.

Mutation combinations confer distinct,
disease-relevant tumor phenotypes in vivo

We next related multimutant genotypes to their
tumor phenotypes in vivo by injecting each
immortalized melanoma model into the dermis

of immunodeficient mice and assessing for

disease-relevant features (Fig. 3A). CBT melano-

cytes were malignant in vivo and formed slow-

growing tumors in xenograft models. Between

67 and 111 days, no primary tumor growth was

detected [Fig. 3, B and C, (n=4)and(n=8),

respectively, black curves; and fig. S8 (n=8)].

However, upon tissue harvest, small nodules

were visible at the injection sites. Histologic

and immunophenotypic evaluation confirmed

the presence of melanoma cells in these nodules

(six of six tumors examined; figs. S9 to S12 and

tables S5 to S7), with half of the nodules also

displaying adjacent features of a congenital

nevus (three of six tumors). Over a longer time

course (≥150 days), a small tumor (≤14 mm^3 )

Hodiset al.,Science 376 , eabi8175 (2022) 29 April 2022 4of14

B

A

D

C

UMAP 1

UMAP 2

WT

C

CB

CBT

CBT3

CBTA

Program 1

Program 2

Interferon / p53

Program 1

Melanocyte

Program 6

S Phase

Program 7

G2 / Mitosis

Program 2 Program 3

Program 3

EMT

Program 5 Program 6 Program 7

Program 4

Program 4

Myc / mTORC1

Program 5

Myc / mTORC1 / Ox-Phos

CBTP

CBTP3

CBTPA

WT

CB

CBT

CBTP

CBTP3

CBTPA

CBT3

CBTA

C

0

25

50

75

100

Expressing

cells (%)

–2

–1

0

1

2

Expression level

(z-norm pseudo-bulk

log2(TP10K))

6. Fatty acid metabolism


7. Unfolded protein response*


8. Ox-Phos**


9. Myc targets v2**


10. mTORC1 signaling**


11. Myc targets v1
    4. Spermatogenesis
    3. G2M checkpoint
    2. Myc targets v1**
    
    
    1. E2F targets
       3. Mitotic spindle
       4. Spermatogenesis
       2. E2F targets**
       
       
       1. G2M checkpoint**
       
       
       
       
    
    
    
    

0.5

0.4

0.3

0.2

0.1

0.0

0.6

0.4

0.2

0.0

0.6

0.8

0.4

0.2

0.0

0.6

0.8

0.4

0.2

0.0

0.6

0.8

1.0

0.4

0.2

0.0

0.6

0.8

1.0

0.4

0.2

0.0

0.6

0.8

0.4

0.2

0.0

Usage

Gene setenrichment

cNMF

UMAP 1 UMAP 1 UMAP 1

UMAP 1

UMAP 2

UMAP 2

WT

C

CBCBT

CBT3

CBTA

CBTPCBTP3

CBTPA

1.0

0.8

0.6

0.4

0.2

0.0

Program

usage

5. Coagulation


6. Adipogenesis


7. Ox-Phos


8. p53 pathway*


9. Interferon response*

Melanocyte-related 1. Interferon response**

(by manual review)

All: adj. p < 5 × 10–4; *adj. p < 1 × 10–6; **adj. p < 1 × 10–12

5. Apical junction


6. Myogenesis


7. TGF signaling


8. Inflammatory response


9. TNF signaling via NFB*


10. Coagulation*


11. UV response down*


12. Epithelial-mesenchymal
    transition (EMT)*
    6. IL2 STAT5 signaling
    7. TNF signaling via NFB
    5. Wnt -catenin signaling
    4. ROS pathway
    3. mTORC1 signaling
    2. Myc targets v2*
    
    
    1. Myc targets v1**
    
    
    
    

UBE2C

KPNA2

ASPM

TOP2A

DLGAP5

PLK1

CENPF

TUBB4B

AURKA

NUSAP1

CENPE

ARL6IP1

TPX2

CCNB1

GTSE1

DUT

PCNA

RRM2TYMS

PCLAF

CLSPN

ATAD2

SIVA1

HIST1H4C

HIST1H1A

FEN1

CENPU

CARHSP1

HELLS

HIST1H1D

ACSL3

WFDC1

HSP90AB1

MT-ND5

SLC27A3

PGK1

GPR137B

ZFYVE16

BCL2A1

SMS

YWHAZ

CAPG

SDCBP

SPON2

FABP5

IL24

MME

MYOZ2

ATOX1

USP53

ABHD2

CSRP2

HERPUD1

STXBP1

SMIM3

NQO1

CTSB

RHOBTB3

PRDX1

ACSL1

ANXA1

SERPINE2

PRSS23

TIMP3

FN1VIM

MYL12B

PMEPA1

S100A10

TUBA1A

DAG1

RCAN1

TMSB10

LGALS1

ANXA2

LY6E

IFI6

LY96

B2M

HLA-B

HLA-C

SPP1

ISG15

SH3BGRL3

CYBA

APOE

FKBP1A

C11orf96

ATP6V1F

PDCD6

DCT

RAB32

ZNF106TYRP1

TRPM1

LIMA1

MYO10

TFAP2A

ASAH1

TBC1D16

RPL19

RPS12

PIP4K2A

TEX41

PEBP1

Label cells with a unique, DNA-barcoded,

cell-surface-protein antibody

Pool all

genotypes

scRNA-Seq Joint analysis

of all genotypes

WT C CB CBT

CBT3 CBTA CBTP CBTP3

CBTPA

Fig. 2. Consecutive mutations produce ordered progression through
expression space and activate shared expression programs.(A) Experimental
overview to profile gene expression in parallel from cells from nine engineered
genotypes with hashed scRNA-seq. (B) Gradual progression of cell states with
genotype. UMAP embedding of melanocyte scRNA-seq profiles (dots) colored and
labeled (boxes) by genotype (legend). Arrows follow the editing tree (as in Fig. 1B).
(C) Expression programs. UMAP embedding as in (B) colored by per-cell relative
usage (color bar) of each of the seven expression programs identified by cNMF.

(D) Programs reflect key processes and vary across genotypes. Top: Distribution

of relative program usages (yaxis) in single cells of each genotype (xaxis,

color legend). Middle: Aggregate (pseudobulk) expression [Z score of expression

(log 2 of transcripts per 10,000 reads, TP10K), color bar] and percent of expressing

single cells (white circles) of the 15 top program-associated genes (rows) per

genotype (columns). Bottom: Ranked lists of gene sets (MSigDB hallmark) ( 83 )

enriched in each program (Mann-WhitneyUtest, false discovery rate (FDR) < 5 × 10−^4 ;

*FDR<10−^6 ; **FDR < 10−^12 ).

RESEARCH | RESEARCH ARTICLE